Abstarct: Abstract The dynamic performance of DC microgrids is highly affected by the variations in power consumption from electric loads and the power generated by renewable energy sources. Additionally, the existence of electric loads with constant power inherent has degraded the dynamic behavior of DC microgrids. Many control strategies to improve the dynamic characteristics of a DC microgrid with constant power loads (CPLs) are recently reported in the literature. One such strategy is Model Predictive Control (MPC). MPC approaches for DC microgrids can be classified into two categories: a) MPC in the grid level, and b) MPC in the converter level. The former is responsible for power distribution, load management, power quality, and economic efficiency, while the latter is responsible for local control of converters to control the output voltage in islanded mode and the output power in grid-connected mode. Recently, Direct Voltage Model Predictive Control (DVMPC) and Direct Power Model Predictive Control (DPMPC) methods have been presented for controlling a boost converter used in DC microgrids. However, the impact of CPLs on the dynamic response of the DVMPC and the effect of long-term prediction on the performance of the DPMPC are not taken into account. To address this issue, this paper proposes a model for automatic prediction of the boost converter in the presence of resistive and constant power loads. Next, two long horizon prediction-baxsed DVMPC and DPMPC strategies are respectively proposed for improving the dynamic behavior of the boost converter in islanded and grid-connected modes. In addition, a move blocking strategy for limiting the computational burden of the proposed strategies is introduced. Simulation results carried out using MATLAB/Simulixnk software show that the proposed strategies successfully improve the dynamic performance of the boost converter in different operating conditions including start-up, reference change, and CPL variation. Finally, the proposed strategies are verified using an experimental setup implemented in TMS320F28335 DSP, and the experimental results confirm the capability of the proposed strategies in enhancing the dynamic performance of the boost converter and, therefore, the DC microgrid.